
%load_ext autoreload
%autoreload 2

# TODO remove
import sys,os
sys.path.append(os.path.expanduser('~/imodels'))
sys.path.append(os.path.expanduser('~/dtreeviz'))

########################################################
# python

import os
import time
import pandas as pd
import numpy as np
import scipy.stats
norm = scipy.stats.norm
import bisect
import warnings

########################################################
# figs (imodels), xgboost, sklearn

import imodels
from imodels import FIGSClassifier
from imodels.tree.viz_utils import extract_sklearn_tree_from_figs

import xgboost as xgb

from sklearn.datasets import make_classification
from sklearn.model_selection import train_test_split
from sklearn import metrics
from sklearn.metrics import roc_curve, precision_recall_curve
from sklearn.inspection import permutation_importance

########################################################
# dtreeviz
# must follow the package README to properly install all dependencies!

from dtreeviz import trees
from dtreeviz.models.sklearn_decision_trees import ShadowSKDTree
from dtreeviz.models.xgb_decision_tree import ShadowXGBDTree
from dtreeviz.colors import mpl_colors

from wand.image import Image
from svglib.svglib import svg2rlg
from reportlab.graphics import renderPDF

########################################################
# skompiler

from skompiler import skompile

########################################################
# plotting

import matplotlib as mpl
import matplotlib.pyplot as plt
import matplotlib.transforms
%matplotlib inline

from plotting import *

########################################################
# set global rnd_seed for reproducibility

rnd_seed = 42
np.random.seed(rnd_seed)

datasets = ['train', 'holdout']

<frozen importlib._bootstrap>:283: DeprecationWarning: the load_module() method is deprecated and slated for removal in Python 3.12; use exec_module() instead


inline=True # plot inline or to pdf
output = './output_figs_demo' # output dir
os.makedirs(output, exist_ok=True)


def save_dtreeviz(viz, m_path, fname, tag='', inline=inline, svg=False, png=False, pdf=True):
    if inline:
        display(viz)
    else:
        if not (svg or png or pdf):
            warnings.warn('Not saving anything!')

        os.makedirs(m_path, exist_ok=True)
        full_path = f'{m_path}/{fname}{tag}'

        # svg
        viz.save(f'{full_path}.svg')

        # pdf via svglib
        if pdf:
            renderPDF.drawToFile(svg2rlg(f'{full_path}.svg'), f'{full_path}.pdf')

        # png via wand / ImageMagick
        if png:
            img = Image(filename=f'{full_path}.svg', resolution=500)
            img.format = 'png'
            img.save(filename=f'{full_path}.png')

        if not svg:
            # clean up svg
            os.remove(f'{full_path}.svg')

        # clean up graphviz dot file (no extension)
        os.remove(full_path)


def save_plt(m_path, fname, tag='', inline=inline):
    plt.tight_layout()
    if inline:
        plt.show()
    else:
        os.makedirs(m_path, exist_ok=True)
        plt.savefig(f'{m_path}/{fname}{tag}.pdf')
        plt.close('all')


mc_params_all = {'n_samples': int(1e5), 'n_classes': 2, 'shuffle': False, 'shift': 0.0, 'scale': 1.0, 'hypercube': True}

mc_params = [
    {'n_features': 20, 'n_informative': 6, 'n_redundant': 4, 'n_repeated': 0,
     'n_clusters_per_class': 5, 'weights': [0.5], 'flip_y': 0.05, 'class_sep': 0.9},
    {'n_features': 10, 'n_informative': 4, 'n_redundant': 2, 'n_repeated': 0,
     'n_clusters_per_class': 2, 'weights': [0.7], 'flip_y': 0.1, 'class_sep': 0.9},
    {'n_features': 5, 'n_informative': 2, 'n_redundant': 2, 'n_repeated': 0,
     'n_clusters_per_class': 2, 'weights': [0.6], 'flip_y': 0.04, 'class_sep': 0.9},
]

X = None
y = None
feat_names = []

for i_mc_param, mc_param in enumerate(mc_params):
    param = {**mc_params_all, **mc_param, 'random_state': rnd_seed+i_mc_param}
    X_i, y_i = make_classification(**param)
    if X is None:
        X = X_i
    else:
        X = np.concatenate([X, X_i], axis=1)
    if y is None:
        y = y_i
    else:
        y = np.logical_and(y, y_i).astype(int)
    feat_names += [f'x_{i_mc_param}_{_}' for _ in range(X_i.shape[1])]
    del X_i; del y_i;


X_trainVal, X_holdout, y_trainVal, y_holdout = train_test_split(X, y, test_size=0.15, random_state=rnd_seed, stratify=y)
del X; del y;

X_train, X_val, y_train, y_val = train_test_split(X_trainVal, y_trainVal, test_size=0.2, random_state=rnd_seed, stratify=y_trainVal)
# del X_trainVal; del y_trainVal;


model_figs = FIGSClassifier(max_rules=20)


time_figs_start = time.time()
model_figs.fit(X_trainVal, y_trainVal, feature_names=feat_names);
time_figs_end = time.time()
print(f'FIGS ran in {time_figs_end-time_figs_start:.0f} seconds')

FIGS ran in 15 seconds


def count_splits_figs(model):
    splits = []
    for tree_ in model.trees_:
        node_counter = iter(range(1, int(1e06)))
        def _count_node(node):
            if node.left is None:
                return
            node_id=next(node_counter)
            _count_node(node.left)
            _count_node(node.right)

        _count_node(tree_)
        splits.append(next(node_counter)-1)
    return sum(splits)

n_splits_figs = count_splits_figs(model_figs)


print(f'FIGS used {len(model_figs.trees_)} trees and {n_splits_figs:,} splits')

FIGS used 2 trees and 20 splits


print(model_figs)

> ------------------------------
> FIGS-Fast Interpretable Greedy-Tree Sums:
> 	Predictions are made by summing the "Val" reached by traversing each tree
> ------------------------------
x_2_1 <= 0.048 (Tree #0 root)
	x_2_0 <= -0.608 (split)
		x_1_4 <= 0.018 (split)
			x_0_7 <= 1.400 (split)
				Val: 0.125 (leaf)
				Val: 0.377 (leaf)
			Val: 0.052 (leaf)
		Val: 0.019 (leaf)
	x_0_8 <= 2.363 (split)
		x_0_7 <= 1.870 (split)
			Val: 0.181 (leaf)
			Val: 0.366 (leaf)
		x_1_0 <= 0.339 (split)
			x_2_0 <= -0.046 (split)
				Val: 0.423 (leaf)
				Val: 0.688 (leaf)
			x_1_1 <= 2.087 (split)
				Val: 0.227 (leaf)
				Val: 0.502 (leaf)

	+
x_0_2 <= -0.432 (Tree #1 root)
	Val: -0.040 (leaf)
	x_0_3 <= 0.455 (split)
		Val: -0.015 (leaf)
		x_2_1 <= -0.317 (split)
			Val: 0.013 (leaf)
			x_1_3 <= -0.299 (split)
				x_0_0 <= -0.958 (split)
					x_1_2 <= 0.359 (split)
						Val: -0.021 (leaf)
						Val: 0.437 (leaf)
					x_2_0 <= 0.152 (split)
						Val: 0.414 (leaf)
						Val: 0.214 (leaf)
				x_1_4 <= 0.165 (split)
					x_0_5 <= 0.930 (split)
						x_2_0 <= 0.567 (split)
							Val: 0.469 (leaf)
							Val: 0.063 (leaf)
						Val: -0.027 (leaf)
					x_1_5 <= -0.480 (split)
						Val: 0.162 (leaf)
						Val: -0.151 (leaf)


print(model_figs.print_tree(X_train, y_train))

------------
x_2_1 <= 0.048 9694/68000 (14.26%)
	x_2_0 <= -0.608 1205/41959 (2.87%)
		x_1_4 <= 0.018 1139/17540 (6.49%)
			x_0_7 <= 1.400 693/3569 (19.42%)
				ΔRisk = 0.13 286/2525 (11.33%)
				ΔRisk = 0.38 407/1044 (38.98%)
			ΔRisk = 0.05 446/13971 (3.19%)
		ΔRisk = 0.02 66/24419 (0.27%)
	x_0_8 <= 2.363 8489/26041 (32.6%)
		x_0_7 <= 1.870 4371/19070 (22.92%)
			ΔRisk = 0.18 2504/13936 (17.97%)
			ΔRisk = 0.37 1867/5134 (36.37%)
		x_1_0 <= 0.339 4118/6971 (59.07%)
			x_2_0 <= -0.046 3206/4510 (71.09%)
				ΔRisk = 0.42 540/1055 (51.18%)
				ΔRisk = 0.69 2666/3455 (77.16%)
			x_1_1 <= 2.087 912/2461 (37.06%)
				ΔRisk = 0.23 462/1698 (27.21%)
				ΔRisk = 0.50 450/763 (58.98%)

	+
x_0_2 <= -0.432 9694/68000 (14.26%)
	ΔRisk = -0.04 1458/28479 (5.12%)
	x_0_3 <= 0.455 8236/39521 (20.84%)
		ΔRisk = -0.01 3164/21232 (14.9%)
		x_2_1 <= -0.317 5072/18289 (27.73%)
			ΔRisk = 0.01 354/8539 (4.15%)
			x_1_3 <= -0.299 4718/9750 (48.39%)
				x_0_0 <= -0.958 3026/5107 (59.25%)
					x_1_2 <= 0.359 411/1412 (29.11%)
						ΔRisk = -0.02 196/1117 (17.55%)
						ΔRisk = 0.44 215/295 (72.88%)
					x_2_0 <= 0.152 2615/3695 (70.77%)
						ΔRisk = 0.41 1217/1737 (70.06%)
						ΔRisk = 0.21 1398/1958 (71.4%)
				x_1_4 <= 0.165 1692/4643 (36.44%)
					x_0_5 <= 0.930 823/1661 (49.55%)
						x_2_0 <= 0.567 651/1052 (61.88%)
							ΔRisk = 0.47 541/736 (73.51%)
							ΔRisk = 0.06 110/316 (34.81%)
						ΔRisk = -0.03 172/609 (28.24%)
					x_1_5 <= -0.480 869/2982 (29.14%)
						ΔRisk = 0.16 724/1325 (54.64%)
						ΔRisk = -0.15 145/1657 (8.75%)


if inline:
    model_figs.plot(fig_size=7)


params_default = {'max_depth': 6, 'learning_rate': 0.3, 'gamma': 0.0, 'reg_alpha': 0.0, 'reg_lambda': 1.0}


fixed_setup_params = {
    'max_num_boost_rounds': 500, # maximum number of boosting rounds to run / trees to create
    'xgb_objective': 'binary:logistic', # objective function for binary classification
    'xgb_verbosity': 0, #  The degree of verbosity. Valid values are 0 (silent) - 3 (debug).
    'xgb_n_jobs': -1, # Number of parallel threads used to run XGBoost. -1 makes use of all cores in your system
    'eval_metric': 'auc', # evaluation metric for early stopping
    'early_stopping_rounds': 10, # must see improvement over last num_early_stopping_rounds or will halt
}


fixed_fit_params = {
    'eval_set': [(X_val, y_val)], # data sets to use for early stopping evaluation
    'verbose': False, # even more verbosity control
}


model_xgboost = xgb.XGBClassifier(n_estimators=fixed_setup_params['max_num_boost_rounds'],
                                  objective=fixed_setup_params['xgb_objective'],
                                  verbosity=fixed_setup_params['xgb_verbosity'],
                                  eval_metric=fixed_setup_params['eval_metric'],
                                  early_stopping_rounds=fixed_setup_params['early_stopping_rounds'],
                                  random_state=rnd_seed+3, **params_default)


time_xgboost_start = time.time()
model_xgboost.fit(X_train, y_train, **fixed_fit_params);
model_xgboost.get_booster().feature_names = feat_names
time_xgboost_end = time.time()
print(f'XGBoost ran in {time_xgboost_end-time_xgboost_start:.0f} seconds')

XGBoost ran in 5 seconds


n_splits_xgboost = sum([tree.count('"split"') for tree in model_xgboost.get_booster().get_dump(dump_format='json')[0:model_xgboost.best_ntree_limit]])


print(f'XGBoost used {model_xgboost.best_ntree_limit} trees and {n_splits_xgboost:,} splits')

XGBoost used 58 trees and 2,767 splits


def classifier_metrics(model, model_nname, X_train, y_train, X_holdout, y_holdout, feature_names, do_permutation_importance=True, print_classification_report=False):
    model_metrics = {'nname': model_nname}
    dfp_importance = pd.DataFrame({'feature': feature_names})
    dfp_importance['icolX'] = dfp_importance.index

    for dataset in datasets[::-1]:
        if dataset == 'holdout':
            X = X_holdout
            y = y_holdout
        elif dataset == 'train':
            X = X_train
            y = y_train
        y_pred = model.predict(X)
        # only want positive class prob
        try:
            # use best_iteration for XGBoost
            y_pred_prob = model.predict_proba(X, iteration_range=(0, model.best_iteration+1))[:, 1]
        except:
            y_pred_prob = model.predict_proba(X)[:, 1]

        model_metrics[dataset] = {}
        model_metrics[dataset]['accuracy_score'] = metrics.accuracy_score(y, y_pred)
        model_metrics[dataset]['precision_score'] = metrics.precision_score(y, y_pred, zero_division=0) # zero_division=0 hides divide by zero warnings that come up with LR doesn't converge
        model_metrics[dataset]['recall_score'] = metrics.recall_score(y, y_pred)
        model_metrics[dataset]['f1_score'] = metrics.f1_score(y, y_pred)
        model_metrics[dataset]['roc_auc_score'] = metrics.roc_auc_score(y, y_pred_prob)
        model_metrics[dataset]['average_precision_score'] = metrics.average_precision_score(y, y_pred_prob) # PR ROC AUC
        model_metrics[dataset]['log_loss'] = metrics.log_loss(y, y_pred)
        model_metrics[dataset]['cohen_kappa_score'] = metrics.cohen_kappa_score(y, y_pred)
        # https://scikit-learn.org/stable/modules/generated/sklearn.metrics.confusion_matrix.html
        CM = metrics.confusion_matrix(y, y_pred)
        model_metrics[dataset]['confusion_matrix'] = CM
        model_metrics[dataset]['TN'] = CM[0][0]
        model_metrics[dataset]['FP'] = CM[0][1]
        model_metrics[dataset]['FN'] = CM[1][0]
        model_metrics[dataset]['TP'] = CM[1][1]
        model_metrics[dataset]['TNR'] = CM[0][0] / (CM[0][0] + CM[0][1]) # TN / (TN + FP)
        model_metrics[dataset]['NPV'] = CM[0][0] / (CM[0][0] + CM[1][0]) # TN / (TN + FN)

        model_metrics[dataset]['pop_PPV'] = len(np.where(y == 1)[0]) / len(y) # P / (P + N)

        # model_metrics[dataset]['dfp_y'] = pd.DataFrame( {'y': y, 'y_pred_prob': y_pred_prob, 'y_pred': y_pred} )

        # for LR models
        # model_converged = (model.n_iter_ < model.max_iter)[0]

        if print_classification_report:
            print(f'For {dataset}:')
            print(metrics.classification_report(y, y_pred))

        # ROC Curves
        def get_n_predicted_positive_vs_thr(y_pred_prob, thr):
            y_pred_prob_sorted = sorted(y_pred_prob)
            return [len(y_pred_prob_sorted) - bisect.bisect_left(y_pred_prob_sorted, _thr) for _thr in thr]

        fpr, tpr, thr_of_fpr_tpr = roc_curve(y, y_pred_prob)
        n_predicted_positive_vs_thr_of_fpr_tpr = get_n_predicted_positive_vs_thr(y_pred_prob, thr_of_fpr_tpr)
        dfp_eval_fpr_tpr = pd.DataFrame({'fpr': fpr, 'tpr': tpr, 'thr': thr_of_fpr_tpr, 'n_predicted_positive': n_predicted_positive_vs_thr_of_fpr_tpr}).sort_values(by='thr').reset_index(drop=True)

        precision, recall, thr_of_precision_recall = precision_recall_curve(y, y_pred_prob)
        thr_of_precision_recall = np.insert(thr_of_precision_recall, 0, [0])
        n_predicted_positive_vs_thr_of_precision_recall = get_n_predicted_positive_vs_thr(y_pred_prob, thr_of_precision_recall)
        dfp_eval_precision_recall = pd.DataFrame({'precision': precision, 'recall': recall, 'thr': thr_of_precision_recall, 'n_predicted_positive': n_predicted_positive_vs_thr_of_precision_recall})
        dfp_eval_precision_recall['f1'] = 2*(dfp_eval_precision_recall['precision'] * dfp_eval_precision_recall['recall']) / (dfp_eval_precision_recall['precision'] + dfp_eval_precision_recall['recall'])

        model_metrics[dataset]['dfp_eval_fpr_tpr'] = dfp_eval_fpr_tpr
        model_metrics[dataset]['dfp_eval_precision_recall'] = dfp_eval_precision_recall

        roc_entry = {'name': f'{model_nname.lower()}_{dataset}',
                     'nname': f'{model_nname} ({dataset.title()})',
                     'dfp_eval_fpr_tpr': dfp_eval_fpr_tpr,
                     'dfp_eval_precision_recall': dfp_eval_precision_recall
                    }
        if dataset == 'holdout':
            roc_entry['c'] = 'C2'
            roc_entry['ls'] = '-'
        else:
            roc_entry['c'] = 'black'
            roc_entry['ls'] = ':'

        model_metrics[dataset]['roc_entry'] = roc_entry

        if do_permutation_importance:
            # print('Start do_permutation_importance func')
            # Permutation feature importance
            # slow for thousands of features!
            # https://scikit-learn.org/stable/modules/generated/sklearn.inspection.permutation_importance.html
            # https://scikit-learn.org/stable/modules/permutation_importance.html#permutation-importance
            # https://scikit-learn.org/stable/auto_examples/inspection/plot_permutation_importance.html#sphx-glr-auto-examples-inspection-plot-permutation-importance-py
            # https://scikit-learn.org/stable/auto_examples/inspection/plot_permutation_importance_multicollinear.html#sphx-glr-auto-examples-inspection-plot-permutation-importance-multicollinear-py
            _permutation_importance = permutation_importance(model, X, y, n_repeats=10, random_state=42, n_jobs=-1, scoring='roc_auc')
            # print('End permutation_importance func')

            importance_permutation_mean = _permutation_importance['importances_mean']
            importance_permutation_std = _permutation_importance['importances_std']
            dfp_importance_permutation = pd.DataFrame({f'importance_permutation_{dataset}_mean': importance_permutation_mean, f'importance_permutation_{dataset}_std': importance_permutation_std})
            dfp_importance_permutation['icolX'] = dfp_importance_permutation.index
            dfp_importance_permutation[f'importance_permutation_{dataset}_pct'] = dfp_importance_permutation[f'importance_permutation_{dataset}_mean'].rank(pct=True)
            dfp_importance = pd.merge(dfp_importance, dfp_importance_permutation, on='icolX', how='left')

    # for LR models
    # dfp_coef = pd.DataFrame({'coefficients': model.coef_[0]})
    # dfp_coef['abs_coeff'] = dfp_coef['coefficients'].abs()
    # dfp_coef['icolX'] = dfp_coef.index
    # dfp_importance = pd.merge(dfp_importance, dfp_coef, on='icolX', how='left')

    # TODO
    # Gini impurity importance - a mean decrease in impurity (MDI) importance (both RF and BDT)
    # https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassifier.html#sklearn.ensemble.RandomForestClassifier.feature_importances_
    # importances_gini = model.feature_importances_
    # estimators = model.estimators_
    # importances_gini_std = np.std([tree.feature_importances_ for tree in estimators], axis=0)
    # dfp_importance_gini = pd.DataFrame({'importance_gini': importances_gini, 'importance_gini_std': importances_gini_std})
    # dfp_importance_gini['icolX'] = dfp_importance_gini.index
    # dfp_importance_gini['importance_gini_pct'] = dfp_importance_gini['importance_gini'].rank(pct=True)
    # dfp_importance = pd.merge(dfp_importance, dfp_importance_gini, on='icolX', how='left')

    target_cols_importance = [
        'feature',
        # 'coefficients',
        'importance_permutation_holdout_mean',
        'importance_permutation_holdout_std',
        'importance_permutation_holdout_pct',
        'importance_permutation_train_mean',
        'importance_permutation_train_std',
        'importance_permutation_train_pct',
        # 'importance_gini',
        # 'importance_gini_std',
        # 'importance_gini_pct',
        'icolX',
        # 'abs_coeff',
    ]
    _cols = [_col for _col in target_cols_importance if _col in dfp_importance.columns] + [_col for _col in dfp_importance.columns if _col not in target_cols_importance]
    dfp_importance = dfp_importance[_cols]
    if 'importance_permutation_holdout_mean' in dfp_importance.columns:
        sort_col = 'importance_permutation_holdout_mean'
    elif 'importance_gini' in dfp_importance.columns:
        sort_col = 'importance_gini'
    else:
        sort_col = 'icolX'
    dfp_importance = dfp_importance.sort_values(by=sort_col, ascending=False).reset_index(drop=True)

    dfp_importance = dfp_importance.drop(['icolX'], axis=1)
    # if 'abs_coeff' in dfp_importance.columns:
    #     dfp_importance = dfp_importance.drop(['abs_coeff'], axis=1)

    model_metrics['dfp_importance'] = dfp_importance

    return model_metrics


model_metrics_figs = classifier_metrics(model_figs, 'FIGS', X_trainVal, y_trainVal, X_holdout, y_holdout, feat_names)
model_metrics_xgboost = classifier_metrics(model_xgboost, 'XGBoost', X_train, y_train, X_holdout, y_holdout, feat_names)

metric_rows = []
for model_metrics in [model_metrics_figs, model_metrics_xgboost]:
    roc_entries = []
    for idataset,dataset in enumerate(datasets[::-1]):
        dataset_metrics = {'model': model_metrics['nname'], 'dataset': dataset}
        for k,v in model_metrics[dataset].items():
            if k == 'roc_entry':
                v = v.copy()
                if 0 < len(roc_entries):
                    v['name'] = dataset
                roc_entries.append(v)
            elif k not in ['confusion_matrix', 'dfp_eval_fpr_tpr', 'dfp_eval_precision_recall', 'dfp_y']:
                dataset_metrics[k] = v
        metric_rows.append(dataset_metrics)

    plot_rocs(roc_entries, m_path=f'{output}/roc_curves', rndGuess=False, inverse_log=False, inline=False)
    plot_rocs(roc_entries, m_path=f'{output}/roc_curves', rndGuess=False, inverse_log=False, precision_recall=True,
              pop_PPV=model_metrics['holdout']['pop_PPV'], y_axis_params={'min': -0.05}, inline=False)

dfp_metrics = pd.DataFrame(metric_rows)
dfp_metrics = dfp_metrics.sort_values(by=['model', 'dataset'], ascending=[True, True]).reset_index(drop=True)
display(dfp_metrics)


print(f'XGBoost used {n_splits_xgboost:,} splits vs FIGS {n_splits_figs:,}')
print(f'That is {n_splits_xgboost-n_splits_figs:,}, or {(n_splits_xgboost-n_splits_figs)/n_splits_figs:,.0%}, more splits!')

XGBoost used 2,767 splits vs FIGS 20
That is 2,747, or 13,735%, more splits!


print('FIGS Feature Importances')
_dfp = model_metrics_figs['dfp_importance']
display(_dfp.loc[0 < _dfp['importance_permutation_holdout_mean']])

FIGS Feature Importances


print('XGBoost Feature Importances')
_dfp = model_metrics_xgboost['dfp_importance']
display(_dfp.loc[0 < _dfp['importance_permutation_holdout_mean']])

XGBoost Feature Importances


for dataset in datasets[::-1]:
    roc_entry_figs = model_metrics_figs[dataset]['roc_entry']
    roc_entry_figs['c'] = 'C0'
    roc_entry_figs['ls'] = '--'
    roc_entry_xgboost = model_metrics_xgboost[dataset]['roc_entry']
    roc_entry_xgboost['c'] = 'C1'
    roc_entry_xgboost['ls'] = ':'

    models_for_roc = [roc_entry_figs, roc_entry_xgboost]
    plot_rocs(models_for_roc, m_path=f'{output}/roc_curves', rndGuess=False, inverse_log=False, inline=False)
    plot_rocs(models_for_roc, m_path=f'{output}/roc_curves', rndGuess=False, inverse_log=False, precision_recall=True,
              pop_PPV=model_metrics_figs[dataset]['pop_PPV'], y_axis_params={'min': -0.05}, inline=False)

if inline:
    for fname in ['roc_figs_holdout_xgboost_holdout', 'roc_figs_holdout_train', 'roc_xgboost_holdout_train']:
        img = Image(filename=f'{output}/roc_curves/{fname}.pdf')
        display(img)


color_params = {'classes': mpl_colors, 'hist_bar': 'C0', 'legend_edge': None}
for _ in ['axis_label', 'title', 'legend_title', 'text', 'arrow', 'node_label', 'tick_label', 'leaf_label', 'wedge', 'text_wedge']:
    color_params[_] = 'black'
dtv_params_gen = {'colors': color_params, 'fontname': 'Arial'} # 'fontname': 'Helvetica'
dtv_params = {'leaf_plot_type': 'barh', 'all_axes_spines': False, 'label_fontsize': 10, 'colors': dtv_params_gen['colors'], 'fontname': dtv_params_gen['fontname']}


x_example = X_train[13]
feature_to_look_at_in_detail = 'x_1_1'


pd.DataFrame([{col: value for col,value in zip(feat_names, x_example)}])


dt_figs_0 = extract_sklearn_tree_from_figs(model_figs, tree_num=0, n_classes=2)
shadow_figs_0 = ShadowSKDTree(dt_figs_0, X_train, y_train, feat_names, 'y', [0, 1])

dt_figs_1 = extract_sklearn_tree_from_figs(model_figs, tree_num=1, n_classes=2)
shadow_figs_1 = ShadowSKDTree(dt_figs_1, X_train, y_train, feat_names, 'y', [0, 1])


viz = trees.dtreeviz(shadow_figs_0, **dtv_params)
save_dtreeviz(viz, output, 'dtreeviz_figs_0')


viz = trees.dtreeviz(shadow_figs_1, **dtv_params)
save_dtreeviz(viz, output, 'dtreeviz_figs_1')


viz = trees.dtreeviz(shadow_figs_0, **dtv_params, fancy=False, show_node_labels=True)
save_dtreeviz(viz, output, 'dtreeviz_text_figs_0')


viz = trees.dtreeviz(shadow_figs_1, **dtv_params, fancy=False, show_node_labels=True)
save_dtreeviz(viz, output, 'dtreeviz_text_figs_1')


print(trees.explain_prediction_path(shadow_figs_0, x=x_example, explanation_type='plain_english'))

0.02 <= x_1_4 
x_2_0 < -0.61
x_2_1 < 0.05


viz = trees.dtreeviz(shadow_figs_0, **dtv_params, X=x_example)
save_dtreeviz(viz, output, 'dtreeviz_pred_path_figs_0')


print(trees.explain_prediction_path(shadow_figs_1, x=x_example, explanation_type='plain_english'))

-0.43 <= x_0_2 
x_0_3 < 0.46


viz = trees.dtreeviz(shadow_figs_1, **dtv_params, X=x_example)
save_dtreeviz(viz, output, 'dtreeviz_pred_path_figs_1')


trees.ctreeviz_leaf_samples(shadow_figs_0, **dtv_params_gen)
save_plt(output, 'ctreeviz_leaf_samples_figs_0')


trees.ctreeviz_leaf_samples(shadow_figs_1, **dtv_params_gen)
save_plt(output, 'ctreeviz_leaf_samples_figs_1')


trees.viz_leaf_criterion(shadow_figs_0, display_type='plot', **dtv_params_gen)
save_plt(output, 'viz_leaf_criterion_figs_0')


trees.viz_leaf_criterion(shadow_figs_0, display_type='hist', **dtv_params_gen)
save_plt(output, 'viz_leaf_criterion_hist_figs_0')


trees.viz_leaf_criterion(shadow_figs_1, display_type='plot', **dtv_params_gen)
save_plt(output, 'viz_leaf_criterion_figs_1')


trees.viz_leaf_criterion(shadow_figs_1, display_type='hist', **dtv_params_gen)
save_plt(output, 'viz_leaf_criterion_hist_figs_1')


trees.ctreeviz_univar(shadow_figs_0, feature_name=feature_to_look_at_in_detail, **dtv_params_gen, gtype = 'barstacked', show={'legend', 'splits', 'axis'})
save_plt(output, 'ctreeviz_univar_figs_0')


trees.describe_node_sample(shadow_figs_0, 18)


shadow_xgboost_0 = ShadowXGBDTree(model_xgboost, 0, X_train, y_train, feat_names, 'y', [0, 1])


viz = trees.dtreeviz(shadow_xgboost_0, **dtv_params)
save_dtreeviz(viz, output, 'dtreeviz_xgboost_0')


viz = trees.dtreeviz(shadow_xgboost_0, **dtv_params, fancy=False, show_node_labels=True)
save_dtreeviz(viz, output, 'dtreeviz_text_xgboost_0')


print(trees.explain_prediction_path(shadow_xgboost_0, x=x_example, explanation_type='plain_english'))

-0.92 <= x_1_3 
0.02 <= x_1_4 
x_1_5 < 3.17
x_2_0 < -0.65
x_2_1 < -0.22


viz = trees.dtreeviz(shadow_xgboost_0, **dtv_params, X=x_example)
save_dtreeviz(viz, output, 'dtreeviz_pred_path_xgboost_0')


trees.ctreeviz_leaf_samples(shadow_xgboost_0, **dtv_params_gen, label_all_leafs=False)
save_plt(output, 'ctreeviz_leaf_samples_xgboost_0')


trees.ctreeviz_univar(shadow_xgboost_0, feature_name=feature_to_look_at_in_detail, **dtv_params_gen, gtype = 'barstacked', show={'legend', 'splits', 'axis'})
save_plt(output, 'ctreeviz_univar_xgboost_0')


trees.describe_node_sample(shadow_xgboost_0, 42)


expr_figs_0 = skompile(dt_figs_0.predict_proba, feat_names)


print(expr_figs_0.to('sqlalchemy/sqlite', component=1, assign_to='tree_0'))

SELECT CASE WHEN (x_2_1 <= 0.047700537368655205) THEN CASE WHEN (x_2_0 <= -0.6084129214286804) THEN CASE WHEN (x_1_4 <= 0.017502836883068085) THEN CASE WHEN (x_0_7 <= 1.3998010754585266) THEN 0.11379529561347743 ELSE 0.38880248833592534 END ELSE 0.032956362805574596 END ELSE 0.0026475779564620514 END ELSE CASE WHEN (x_0_8 <= 2.363405704498291) THEN CASE WHEN (x_0_7 <= 1.8703043460845947) THEN 0.1812694032424974 ELSE 0.36647904940587867 END ELSE CASE WHEN (x_1_0 <= 0.3394070416688919) THEN CASE WHEN (x_2_0 <= -0.04638334736227989) THEN 0.5054517133956387 ELSE 0.7712753555607368 END ELSE CASE WHEN (x_1_1 <= 2.0868066549301147) THEN 0.27399165507649514 ELSE 0.5887949260042283 END END END END AS tree_0 
FROM data


print(expr_figs_0.to('python/code'))

((((np.array([0.8862047043865225, 0.11379529561347743]) if x_0_7 <= 
    1.3998010754585266 else np.array([0.6111975116640747, 
    0.38880248833592534])) if x_1_4 <= 0.017502836883068085 else np.
    array([0.9670436371944254, 0.032956362805574596])) if x_2_0 <= 
    -0.6084129214286804 else np.array([0.9973524220435379, 
    0.0026475779564620514])) if x_2_1 <= 0.047700537368655205 else (np.
    array([0.8187305967575026, 0.1812694032424974]) if x_0_7 <= 
    1.8703043460845947 else np.array([0.6335209505941213, 
    0.36647904940587867])) if x_0_8 <= 2.363405704498291 else (np.array
    ([0.4945482866043614, 0.5054517133956387]) if x_2_0 <= 
    -0.04638334736227989 else np.array([0.22872464443926324, 
    0.7712753555607368])) if x_1_0 <= 0.3394070416688919 else np.array(
    [0.7260083449235049, 0.27399165507649514]) if x_1_1 <= 
    2.0868066549301147 else np.array([0.41120507399577166, 
    0.5887949260042283]))


expr_figs_1 = skompile(dt_figs_1.predict_proba, feat_names)


print(expr_figs_1.to('sqlalchemy/sqlite', component=1, assign_to='tree_1'))

SELECT CASE WHEN (x_0_2 <= -0.43244342505931854) THEN 0.050912259987068116 ELSE CASE WHEN (x_0_3 <= 0.4551866203546524) THEN 0.1490697322644078 ELSE CASE WHEN (x_2_1 <= -0.31738142669200897) THEN 0.04208863699711976 ELSE CASE WHEN (x_1_3 <= -0.2993980050086975) THEN CASE WHEN (x_0_0 <= -0.9580468833446503) THEN CASE WHEN (x_1_2 <= 0.35873670876026154) THEN 0.18310858765081617 ELSE 0.7016574585635359 END ELSE CASE WHEN (x_2_0 <= 0.15202737599611282) THEN 0.7007730786721237 ELSE 0.7108632796365139 END END ELSE CASE WHEN (x_1_4 <= 0.16462678462266922) THEN CASE WHEN (x_0_5 <= 0.9298670291900635) THEN CASE WHEN (x_2_0 <= 0.5672585964202881) THEN 0.746875 ELSE 0.3492462311557789 END ELSE 0.2857142857142857 END ELSE CASE WHEN (x_1_5 <= -0.47995996475219727) THEN 0.5458461538461539 ELSE 0.0855640535372849 END END END END END END AS tree_1 
FROM data


print(expr_figs_1.to('python/code'))

(np.array([0.9490877400129318, 0.050912259987068116]) if x_0_2 <= 
    -0.43244342505931854 else np.array([0.8509302677355922, 
    0.1490697322644078]) if x_0_3 <= 0.4551866203546524 else np.array([
    0.9579113630028803, 0.04208863699711976]) if x_2_1 <= 
    -0.31738142669200897 else ((np.array([0.8168914123491838, 
    0.18310858765081617]) if x_1_2 <= 0.35873670876026154 else np.array
    ([0.2983425414364641, 0.7016574585635359])) if x_0_0 <= 
    -0.9580468833446503 else np.array([0.2992269213278763, 
    0.7007730786721237]) if x_2_0 <= 0.15202737599611282 else np.array(
    [0.28913672036348614, 0.7108632796365139])) if x_1_3 <= 
    -0.2993980050086975 else ((np.array([0.253125, 0.746875]) if x_2_0 <=
    0.5672585964202881 else np.array([0.6507537688442211, 
    0.3492462311557789])) if x_0_5 <= 0.9298670291900635 else np.array(
    [0.7142857142857143, 0.2857142857142857])) if x_1_4 <= 
    0.16462678462266922 else np.array([0.45415384615384613, 
    0.5458461538461539]) if x_1_5 <= -0.47995996475219727 else np.array
    ([0.9144359464627151, 0.0855640535372849]))


print(model_xgboost.get_booster()[0].get_dump(dump_format='text')[0])

0:[x_2_1<0.0476587564] yes=1,no=2,missing=1
	1:[x_2_0<-0.645031273] yes=3,no=4,missing=3
		3:[x_1_4<0.0179839358] yes=7,no=8,missing=7
			7:[x_0_2<0.557516992] yes=15,no=16,missing=15
				15:[x_0_7<1.61057043] yes=29,no=30,missing=29
					29:[x_0_5<0.0603608899] yes=53,no=54,missing=53
						53:leaf=-0.451428592
						54:leaf=-0.546934724
					30:[x_2_1<-1.1075592] yes=55,no=56,missing=55
						55:leaf=-0.499441355
						56:leaf=-0.160159379
				16:[x_0_0<-0.517694712] yes=31,no=32,missing=31
					31:[x_0_7<2.26842213] yes=57,no=58,missing=57
						57:leaf=-0.453658551
						58:leaf=0.0285714306
					32:[x_1_0<-0.46045959] yes=59,no=60,missing=59
						59:leaf=-0.272289157
						60:leaf=0.276237637
			8:[x_1_3<-0.922470212] yes=17,no=18,missing=17
				17:[x_1_0<-0.379907161] yes=33,no=34,missing=33
					33:[x_0_3<0.779166937] yes=61,no=62,missing=61
						61:leaf=-0.151401877
						62:leaf=0.400000036
					34:[x_0_2<0.0654732138] yes=63,no=64,missing=63
						63:leaf=-0.520000041
						64:leaf=-0.24000001
				18:[x_2_1<-0.217023164] yes=35,no=36,missing=35
					35:[x_1_5<3.1683383] yes=65,no=66,missing=65
						65:leaf=-0.592393219
						66:leaf=-0.277611941
					36:[x_1_5<-0.842451215] yes=67,no=68,missing=67
						67:leaf=-0.245595872
						68:leaf=-0.522289157
		4:[x_0_11<3.6412077] yes=9,no=10,missing=9
			9:[x_1_0<-3.16866493] yes=19,no=20,missing=19
				19:leaf=-0.150000006
				20:leaf=-0.59614414
			10:leaf=-0.150000006
	2:[x_0_8<2.36315727] yes=5,no=6,missing=5
		5:[x_0_7<1.87101865] yes=11,no=12,missing=11
			11:[x_0_2<-0.327792406] yes=21,no=22,missing=21
				21:[x_0_5<3.19854832] yes=37,no=38,missing=37
					37:[x_1_0<-0.866449833] yes=69,no=70,missing=69
						69:leaf=-0.361528665
						70:leaf=-0.50694263
					38:[x_0_4<-2.11595488] yes=71,no=72,missing=71
						71:leaf=-0.517241418
						72:leaf=0.329577476
				22:[x_0_3<-0.0131379161] yes=39,no=40,missing=39
					39:[x_0_5<2.10414362] yes=73,no=74,missing=73
						73:leaf=-0.4675951
						74:leaf=0.247058839
					40:[x_1_0<-0.741541505] yes=75,no=76,missing=75
						75:leaf=0.0857142955
						76:leaf=-0.260613739
			12:[x_1_1<0.0114513319] yes=23,no=24,missing=23
				23:[x_0_8<-2.22477674] yes=41,no=42,missing=41
					41:[x_0_5<0.787687898] yes=77,no=78,missing=77
						77:leaf=-0.529608965
						78:leaf=0.101538472
					42:[x_1_2<0.552866995] yes=79,no=80,missing=79
						79:leaf=0.201496735
						80:leaf=-0.171577126
				24:[x_1_0<0.651979804] yes=43,no=44,missing=43
					43:[x_1_3<-1.18832505] yes=81,no=82,missing=81
						81:leaf=0.0360975601
						82:leaf=-0.187583163
					44:[x_1_1<2.36274719] yes=83,no=84,missing=83
						83:leaf=-0.548837245
						84:leaf=-0.328813583
		6:[x_1_0<-0.0499799103] yes=13,no=14,missing=13
			13:[x_2_0<-0.0462705716] yes=25,no=26,missing=25
				25:[x_2_1<0.391258895] yes=45,no=46,missing=45
					45:[x_0_8<4.06781816] yes=85,no=86,missing=85
						85:leaf=0.114396892
						86:leaf=0.397260278
					46:[x_0_5<1.06533742] yes=87,no=88,missing=87
						87:leaf=-0.0185022037
						88:leaf=-0.529411793
				26:[x_0_3<0.528353095] yes=47,no=48,missing=47
					47:[x_0_2<2.35224748] yes=89,no=90,missing=89
						89:leaf=0.273134351
						90:leaf=-0.195918366
					48:[x_2_1<0.369207084] yes=91,no=92,missing=91
						91:leaf=-0.375
						92:leaf=0.433992118
			14:[x_1_1<1.66213548] yes=27,no=28,missing=27
				27:[x_0_9<-1.06273139] yes=49,no=50,missing=49
					49:[x_1_0<0.350768745] yes=93,no=94,missing=93
						93:leaf=-0.164705902
						94:leaf=-0.516577542
					50:[x_1_4<0.419323266] yes=95,no=96,missing=95
						95:leaf=0.29815951
						96:leaf=-0.435849071
				28:[x_2_0<-0.396020621] yes=51,no=52,missing=51
					51:[x_2_1<0.192522258] yes=97,no=98,missing=97
						97:leaf=0.0711864457
						98:leaf=-0.556097567
					52:[x_0_3<1.26900887] yes=99,no=100,missing=99
						99:leaf=0.0314647369
						100:leaf=0.347368419

	model	dataset	accuracy_score	precision_score	recall_score	f1_score	roc_auc_score	average_precision_score	log_loss	cohen_kappa_score	TN	FP	FN	TP	TNR	NPV	pop_PPV
0	FIGS	holdout	0.898067	0.712491	0.477549	0.571829	0.911709	0.634618	3.520675	0.516522	12450	412	1117	1021	0.967968	0.917668	0.142533
1	FIGS	train	0.901729	0.733416	0.488117	0.586137	0.914569	0.650357	3.394166	0.532920	70732	2150	6203	5915	0.970500	0.919373	0.142565
2	XGBoost	holdout	0.946467	0.845676	0.763798	0.802654	0.974239	0.844879	1.848992	0.771780	12564	298	505	1633	0.976831	0.961359	0.142533
3	XGBoost	train	0.968147	0.904470	0.868269	0.886000	0.990188	0.946239	1.100172	0.867496	57417	889	1277	8417	0.984753	0.978243	0.142559

	feature	importance_permutation_holdout_mean	importance_permutation_holdout_std	importance_permutation_holdout_pct	importance_permutation_train_mean	importance_permutation_train_std	importance_permutation_train_pct
0	x_2_1	0.281633	0.005927	1.000000	0.283036	0.001755	1.000000
1	x_1_4	0.026162	0.002426	0.971429	0.029008	0.000799	0.942857
2	x_2_0	0.026032	0.001159	0.942857	0.029866	0.000498	0.971429
3	x_0_2	0.024550	0.001272	0.914286	0.023949	0.000588	0.914286
4	x_1_3	0.020139	0.001193	0.885714	0.018275	0.000573	0.857143
5	x_0_8	0.019467	0.001028	0.857143	0.022575	0.000350	0.885714
6	x_0_3	0.014895	0.001420	0.828571	0.014720	0.000550	0.800000
7	x_0_7	0.013435	0.000675	0.800000	0.015834	0.000490	0.828571
8	x_1_5	0.008601	0.000907	0.771429	0.007197	0.000228	0.742857
9	x_1_0	0.007383	0.000489	0.742857	0.008384	0.000237	0.771429
10	x_0_0	0.004723	0.000647	0.714286	0.005603	0.000208	0.714286
11	x_0_5	0.002835	0.000406	0.685714	0.003067	0.000109	0.685714
12	x_1_2	0.002199	0.000611	0.657143	0.003011	0.000198	0.657143
13	x_1_1	0.002138	0.000260	0.628571	0.001666	0.000114	0.628571

	feature	importance_permutation_holdout_mean	importance_permutation_holdout_std	importance_permutation_holdout_pct	importance_permutation_train_mean	importance_permutation_train_std	importance_permutation_train_pct
0	x_2_1	0.242568	0.004872	1.000000	0.234184	0.002396	1.000000
1	x_2_0	0.041755	0.001616	0.971429	0.038610	0.000565	0.971429
2	x_0_5	0.018473	0.000677	0.942857	0.015805	0.000181	0.942857
3	x_1_4	0.015454	0.001070	0.914286	0.013442	0.000286	0.914286
4	x_0_3	0.015285	0.000694	0.885714	0.013131	0.000385	0.857143
5	x_0_2	0.014121	0.000381	0.857143	0.013272	0.000307	0.885714
6	x_0_4	0.009607	0.000705	0.828571	0.009007	0.000238	0.828571
7	x_1_3	0.007532	0.000530	0.800000	0.005411	0.000126	0.771429
8	x_1_5	0.006877	0.000311	0.771429	0.004939	0.000151	0.742857
9	x_0_8	0.005486	0.000396	0.742857	0.005837	0.000167	0.800000
10	x_1_0	0.005371	0.000497	0.714286	0.004837	0.000110	0.714286
11	x_1_1	0.005168	0.000250	0.685714	0.003612	0.000097	0.685714
12	x_0_1	0.003444	0.000395	0.657143	0.003306	0.000082	0.657143
13	x_1_2	0.003442	0.000324	0.628571	0.002383	0.000100	0.600000
14	x_0_7	0.003063	0.000347	0.600000	0.002992	0.000135	0.628571
15	x_0_0	0.002395	0.000263	0.571429	0.002379	0.000123	0.571429
16	x_0_6	0.001672	0.000133	0.542857	0.001782	0.000063	0.542857
17	x_0_9	0.001255	0.000194	0.514286	0.001440	0.000042	0.514286
18	x_2_3	0.000228	0.000112	0.485714	0.000621	0.000034	0.285714
19	x_2_2	0.000224	0.000139	0.457143	0.000502	0.000040	0.114286
20	x_0_14	0.000207	0.000094	0.428571	0.000686	0.000034	0.400000
21	x_1_8	0.000142	0.000062	0.400000	0.000627	0.000023	0.314286
22	x_1_7	0.000104	0.000060	0.371429	0.000515	0.000018	0.142857
23	x_0_17	0.000027	0.000045	0.342857	0.000343	0.000016	0.057143
24	x_0_19	0.000026	0.000067	0.314286	0.000643	0.000035	0.342857
25	x_0_12	0.000019	0.000068	0.285714	0.000544	0.000018	0.171429
26	x_0_15	0.000013	0.000061	0.257143	0.000327	0.000014	0.028571

	x_0_0	x_0_1	x_0_2	x_0_3	x_0_4	x_0_5	x_0_6	x_0_7	x_0_8	x_0_9	...	x_1_5	x_1_6	x_1_7	x_1_8	x_1_9	x_2_0	x_2_1	x_2_2	x_2_3	x_2_4
count	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000	...	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000	763.000000
mean	1.767229	1.116154	1.083245	1.477000	-0.810264	0.525841	2.012303	0.930154	4.182811	-1.129694	...	-0.239827	0.030528	0.026965	0.005057	-0.033579	0.695488	0.997064	1.398869	-0.566830	-0.008997
std	1.450773	1.576796	1.409163	1.775271	1.714659	1.638135	1.895840	2.721929	1.494092	1.491544	...	0.873685	1.009608	0.973880	0.998447	0.989816	1.062380	0.614211	1.054637	0.359206	1.055266
min	-2.401616	-5.031597	-3.034274	-4.715743	-6.532610	-4.867419	-5.427759	-6.078812	2.363504	-5.196776	...	-2.917652	-3.505779	-3.444223	-3.031508	-3.004628	-2.175032	0.053406	-0.903441	-2.532628	-3.825267
25%	0.874623	0.174982	-0.004228	0.207505	-1.898567	-0.554186	0.856166	-0.794236	3.023169	-2.161035	...	-0.826657	-0.642174	-0.567451	-0.650796	-0.706208	-0.106078	0.545339	0.582967	-0.790365	-0.683467
50%	1.719770	1.190583	1.095149	1.433716	-0.826681	0.611711	2.174741	0.829511	3.795663	-1.280435	...	-0.244170	0.025880	0.068336	-0.017602	-0.049954	0.692166	0.941336	1.369705	-0.544056	-0.001038
75%	2.714320	2.104298	2.110253	2.650330	0.291295	1.708167	3.301559	2.772666	4.964354	-0.258586	...	0.241018	0.684262	0.672124	0.668012	0.589396	1.440461	1.309825	2.097894	-0.298490	0.677038
max	6.481590	6.010321	5.060752	7.156921	3.925822	5.455501	7.688588	9.963754	11.028387	3.838186	...	5.072575	3.955030	3.148567	3.396995	3.062164	3.559361	4.857449	5.089502	0.084606	3.342587

	x_0_0	x_0_1	x_0_2	x_0_3	x_0_4	x_0_5	x_0_6	x_0_7	x_0_8	x_0_9	...	x_1_5	x_1_6	x_1_7	x_1_8	x_1_9	x_2_0	x_2_1	x_2_2	x_2_3	x_2_4
count	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	...	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000	1638.000000
mean	1.012674	-0.991336	1.155281	-0.432166	-1.683346	0.098674	0.976336	3.442569	0.434084	-0.579965	...	0.933398	0.024972	0.026276	0.054004	-0.000221	-0.099415	1.230998	1.137298	-0.621605	0.028589
std	1.119107	1.469979	1.278005	1.551187	1.555297	1.187498	1.368490	1.280384	1.231001	1.204688	...	1.649190	0.993177	1.016335	1.005863	1.015380	1.089053	0.852878	1.087048	0.448591	1.033237
min	-3.138056	-6.225011	-4.076778	-5.884318	-6.849363	-3.387708	-2.904388	1.871630	-2.224703	-4.493653	...	-3.276613	-3.115242	-3.373070	-3.868328	-3.710716	-1.983176	0.047943	-1.055921	-2.721531	-3.736635
25%	0.245386	-2.017265	0.327914	-1.568224	-2.749472	-0.760154	-0.082223	2.441920	-0.532810	-1.410550	...	-0.274372	-0.687936	-0.671314	-0.622668	-0.726021	-0.879780	0.588146	0.303620	-0.890519	-0.649375
50%	1.037412	-1.012618	1.146609	-0.578757	-1.694313	0.115276	0.844027	3.176878	0.566483	-0.541788	...	0.792990	0.002374	0.028716	0.059855	-0.034026	-0.413216	1.057586	1.036936	-0.557771	0.035026
75%	1.750700	0.068200	1.975192	0.612229	-0.683499	0.920833	1.944449	4.102262	1.479710	0.280212	...	2.028676	0.686250	0.718958	0.701865	0.678744	0.512686	1.688846	1.831311	-0.277202	0.714184
max	4.974638	3.885834	5.427050	4.218194	3.426993	4.517650	6.719807	9.482532	2.360396	3.521284	...	7.036883	2.928519	3.541607	3.183238	3.181252	4.263644	5.238191	5.415948	0.105382	3.349577

Setup¶

Generate Random Data¶

FIGS¶

XGBoost¶

Evaluate¶

Setup¶

Metrics¶

Feature Importances¶

ROC Curves¶

Tree Plots¶

FIGS¶

Trees¶

Split Hists¶

Text¶

Prediction Path¶

Leaf Samples¶

Leaf Criterion¶

Splits in Feature Space¶

Node Sample¶

XGBoost¶

Trees¶

Split Hists¶

Text¶

Prediction Path¶

Leaf Samples¶

Splits in Feature Space¶

Node Sample¶

Tree Functions¶

FIGS¶

XGBoost¶